Depopulation of the Central Andes in the 16th Century [and Comments and Reply

Abstract

ESTIMATES of the pre-conquest population of the Americas have been the subject of debate and controversy for many years, and the extent to which the Indian population declined after the Spanish conquest has equally been hotly disputed. Dobyns (1966) has revived interest in the subject by his analysis of estimates and methods and by his introduction of the concept of the depopulation ratio to compare regional estimates and to reach a new assessment. There would be no point in recapitulating his arguments and his methodological discussion, but one or two comments may be made. Dobyns concludes that a depopulation ratio of 20:1 or 25:1 between pre-conquest totals and the nadir of Indian population, occurring at various dates in different regions, may represent the truth. He arrives at projections of 90,000,000 and 112,500,000 for the Indian population of the Americas before the impact of the Europeans was felt. The nadir of population in the area of the Central Andean civilizations is estimated at 1,500,000 in 1650 or slightly later (Dobyns 1966:415) and the pre-conquest population at 30,000,000 to 37,500,000, five or six times greater than Rowe's (1947:184) estimrate of 6,000,000 for the pre-conquest population of the area approximately equivalent to the coastal and Andean regions of modern Ecuador, Peru, and Bolivia. It has been pointed out with some justice that a general ratio must necessarily conceal many regional variations, a point well made in the discussion of Dobyns' paper by Denevan (1966:429). In particular, Denevan casts doubt on the applicability of a 20:1 depopulation ratio for the Indian cultures of the Central Andes. For this region, Dobyns relied chiefly on figures derived from Rowe for the Rimac and Chincha areas, which suffered, according to Rowe, depopulation of the order of 16:1 and 20:1 between 1525 and 1571. These coastal regions are not, however, representative of the Central Andes as a whole; Rowe has presented much lower rates of depopulation in the central sierra of Peru (see Table 1, p. 454). Rowe assumed an average depopulation ratio of 4: 1, which was applied to the estimated population of 1,500,000 in 1571 to give a pre-conquest population of 6,000,000. The 1571 figures are based on the totals of tributary Indians recorded in the census of Viceroy Toledo, using an arbitrary multiplier of 5: 1 to estimate total population from the recorded total of 311,257 tributary Indians in the coastal and sierra regions of the Central Andes. It is also assumed that the Inca taxpayers represented a similar proportion of the total population, and the same multiplier of 5:1 is used to convert Inca taxpayers to total numbers. The figures for 1525 are based on two kinds of evidence. For Rimac and Chincha, they are based on no more than the numbers of hunukurakas, the Inca officials responsible for 10,000 taxpayers. Thus the fact that Chincha had one such official is taken as evidence that the province had 10,000 taxpayers, and hence 50,000 people. The magnification of error in such a process of multiplication is clearly enormous. It seems entirely possible that the allocation of 10,000 taxpayers to each Inca official was, or must gradually have become, a nominal allocation, perhaps no more representative of the total population than a count of English hundreds at the time of the Norman conquest. It is unfortunate, therefore, that there is no firmer basis than this for estimnating the fall of population in the coastal region. The estimates for the central sierra of Peru are based on slightly more reliable and direct information: for Yauyos in 1582 (Relaciones geograficas de Indias 1965, vol. 1:155),